Neuromorphic Silicon Photonics

We report first observations of a recurrent silicon photonic neural network, in which connections are configured by microring weight banks. A mathematical isomorphism between the silicon photonic circuit and a continuous neural network model is demonstrated through dynamical bifurcation analysis. Exploiting this isomorphism, a simulated 24-node silicon photonic neural network programmed using an existing "neural compiler" to solve a differential system emulation task. A 294-fold acceleration against a conventional benchmark is predicted. We propose modulator-class neurons that, as opposed to laser-class neurons, are compatible with silicon photonic platforms. Neuromorphic silicon photonics could access new regimes of ultrafast information processing for radio, control, and scientific computing.